A method and system for developing fault models from structured text documents, such as service procedures. A service procedure or other structured text document is parsed using diagnostic rules to extract symptoms, failure modes, and correlations. Testing procedures and repair instructions are also parsed to create a fault tree and identify additional symptoms and failure modes. Reachability analysis is then used to find hidden dependencies in the fault tree, thus yielding additional correlations. The resultant symptoms, failure modes, and correlations are then assembled into a fault model, which can be used for real-time fault diagnosis onboard a vehicle, or for off-board diagnosis at service shops.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method for creating a fault model for a hardware or software system, said method comprising: providing a structured text document containing diagnostic information about the hardware or software system; extracting fault model data from the structured text document; parsing test procedures and repair instructions contained in the structured text document to produce a fault tree and to obtain additional fault model data; performing reachability analysis on the fault tree to identify cross-functional dependencies; and assembling the fault model from the fault model data and the cross-functional dependencies.
A method to automatically build a fault model for a hardware or software system involves these steps: First, provide a structured text document containing diagnostic information about the system. Then, extract initial fault model data (symptoms, failure modes, correlations) from this document. Parse test procedures and repair instructions from the same document to create a fault tree and extract additional fault model data. Next, perform reachability analysis on the fault tree to find hidden cross-functional dependencies between system components. Finally, combine the initial fault model data, additional fault model data, and cross-functional dependencies into a complete fault model.
2. The method of claim 1 wherein extracting fault model data includes using diagnostic rules to extract the fault model data from diagnostic fault information and scan tool tables contained in the structured text document.
The method for creating a fault model as described above, where extracting fault model data from a structured text document is done by applying diagnostic rules to diagnostic fault information and scan tool tables found within the document. This means automatically identifying symptoms, failure modes, and their relationships by analyzing structured diagnostic text and scan tool data, such as parameter IDs and their expected ranges, using predetermined logic rules.
3. The method of claim 1 wherein parsing test procedures and repair instructions includes identifying a test for a first symptom, and identifying a failure mode and a second symptom which may be deduced from the test.
The method for creating a fault model as described above, where parsing test procedures and repair instructions to build the fault tree involves identifying a test for a specific symptom. The process also identifies a failure mode that the test can diagnose and another symptom that can be deduced from the test's results. This allows for linking tests to potential failures and revealing connections between different symptoms through test outcomes.
4. The method of claim 1 wherein performing reachability analysis on the fault tree includes determining where failure modes in the fault tree are cross-functionally dependent on more than one symptom.
The method for creating a fault model as described above, where the reachability analysis on the fault tree determines instances where a failure mode within the tree is cross-functionally dependent on multiple symptoms. This identifies scenarios where a single failure can manifest in different ways, affecting multiple parts of the system and producing different symptoms depending on the operational context.
5. The method of claim 1 wherein the fault model data includes symptoms, failure modes, and correlation values.
The method for creating a fault model as described above, where the fault model data that is extracted includes symptoms (observable indicators of a problem), failure modes (the specific ways a system can fail), and correlation values (quantifying the relationship between symptoms and failure modes). These three elements form the core components of the fault model, allowing for reasoning about the cause of issues based on observed symptoms.
6. The method of claim 5 wherein the symptoms include Diagnostic Trouble Code (DTC) symptoms and non-DTC symptoms.
The method for creating a fault model as described above, where the symptoms included in the fault model data are divided into Diagnostic Trouble Code (DTC) symptoms and non-DTC symptoms. DTC symptoms are standardized error codes reported by the system, while non-DTC symptoms are other observable indicators of a problem that aren't represented by standard codes, such as unusual noises or performance issues.
7. The method of claim 5 wherein assembling the fault model includes creating rows of the failure modes, creating columns of the symptoms, and placing the correlation values in intersections of the rows and the columns.
The method for creating a fault model as described above, where assembling the fault model from the data involves creating a table where each row represents a specific failure mode, each column represents a symptom, and the cell at the intersection of a row and column contains a correlation value. This table visually represents the relationships between failures and symptoms and the strength of those relationships.
8. The method of claim 7 further comprising using the cross-functional dependencies to define additional correlation values in the fault model.
The method for creating a fault model as described above, also includes using the cross-functional dependencies identified during reachability analysis to define additional correlation values in the fault model. These correlation values reflect the complex relationships where one failure mode can lead to multiple symptoms across different system components, thus enhancing the accuracy and comprehensiveness of the fault model.
9. The method of claim 1 wherein the hardware or software system is a vehicle or a vehicle sub-system.
The method for creating a fault model as described above, where the hardware or software system being modeled is specifically a vehicle or a vehicle sub-system. This means the fault model is designed to diagnose problems in cars, trucks, or specific components within them like the engine control unit (ECU) or the braking system.
10. The method of claim 9 wherein the structured text document is a service procedure document for the vehicle or the vehicle sub-system.
The method for creating a fault model as described above, where the hardware or software system is a vehicle or vehicle subsystem, and the structured text document containing diagnostic information is a service procedure document specifically for that vehicle or subsystem. This utilizes existing repair manuals and diagnostic guides to automatically build the fault model.
11. A method for creating a fault model for a vehicle or a vehicle sub-system, said method comprising: providing a service procedure document containing diagnostic information about the vehicle or the vehicle sub-system; using diagnostic rules to extract fault model data from diagnostic fault information and scan tool tables contained in the service procedure document; parsing testing procedures and repair instructions contained in the service procedure document to produce a fault tree and to obtain additional fault model data; performing reachability analysis on the fault tree to identify cross-functional dependencies, where the reachability analysis includes determining where failure modes in the fault tree are cross-functionally dependent on more than one symptom; and assembling the fault model from the fault model data and the cross-functional dependencies.
A method for creating a fault model for a vehicle or vehicle sub-system involves: providing a service procedure document with diagnostic information; using diagnostic rules to extract initial fault model data (symptoms, failure modes, correlations) from diagnostic fault information and scan tool tables in the document; parsing testing procedures and repair instructions to create a fault tree and extract more fault model data; performing reachability analysis on the fault tree to identify cross-functional dependencies, determining where failure modes depend on more than one symptom; and finally assembling the fault model using all the fault model data and cross-functional dependencies.
12. The method of claim 11 wherein the fault model data includes symptoms, failure modes, and correlation values.
The method for creating a fault model for a vehicle or vehicle sub-system, as described above, where the fault model data extracted and used includes symptoms (observable indicators of a problem), failure modes (the specific ways a system can fail), and correlation values (quantifying the relationship between symptoms and failure modes). These form the core components of the fault model, allowing for diagnosis based on observed symptoms.
13. The method of claim 12 wherein assembling the fault model includes creating rows of the failure modes, creating columns of the symptoms, and placing the correlation values in intersections of the rows and the columns.
The method for creating a fault model for a vehicle or vehicle sub-system, as described above, where assembling the fault model involves creating a table where each row represents a failure mode, each column represents a symptom, and the cell at the intersection of a row and column contains a correlation value. This table represents relationships between failures and symptoms and the strength of those relationships.
14. The method of claim 11 further comprising using the fault model for fault diagnosis in connection with the vehicle or the vehicle sub-system.
This invention relates to fault diagnosis in vehicles or vehicle sub-systems using a fault model. The technology addresses the challenge of accurately identifying and diagnosing faults in complex vehicle systems, which is critical for maintaining performance, safety, and reliability. The method involves generating a fault model that represents potential faults and their effects on the vehicle or sub-system. This model is then used to analyze real-time or recorded operational data to detect and diagnose faults. The fault model may include relationships between fault symptoms, root causes, and system responses, enabling precise fault identification. The method can be applied to various vehicle sub-systems, such as powertrain, braking, or electrical systems, to improve diagnostic accuracy and reduce false positives. By leveraging the fault model, the system can provide actionable insights for maintenance, repairs, or system adjustments, enhancing overall vehicle reliability and reducing downtime. The approach may also integrate with existing diagnostic tools or vehicle control systems to streamline fault detection and resolution processes.
15. A system for creating a fault model, said system comprising: means for providing a structured text document containing diagnostic information about a hardware or software system; means for extracting fault model data from the structured text document; means for parsing test procedures and repair instructions contained in the structured text document to produce a fault tree and to obtain additional fault model data; means for performing reachability analysis on the fault tree to identify cross-functional dependencies; and means for assembling the fault model from the fault model data and the cross-functional dependencies.
A system for creating a fault model includes the following components: a module that provides a structured text document containing diagnostic information about a hardware or software system; a module that extracts fault model data from the document; a module that parses test procedures and repair instructions from the document to create a fault tree and get more fault model data; a module that performs reachability analysis on the fault tree to find cross-functional dependencies; and a module that assembles the fault model from the data and dependencies. These modules represent the different stages of the fault model creation process.
16. The system of claim 15 wherein the means for extracting fault model data uses diagnostic rules to extract the fault model data from diagnostic fault information and scan tool tables contained in the structured text document.
The system for creating a fault model as described above, where the module that extracts fault model data uses diagnostic rules to extract the data from diagnostic fault information and scan tool tables found within the structured text document. This automated process mimics how an experienced technician would analyze repair manuals to identify symptoms, failure modes, and their relationships.
17. The system of claim 15 wherein the means for parsing test procedures and repair instructions identifies a test for a first symptom, and identifies a failure mode and a second symptom which may be deduced from the test.
The system for creating a fault model as described above, where the module that parses test procedures and repair instructions identifies a test for a first symptom, and identifies a failure mode and a second symptom which may be deduced from the test. This module automatically analyzes repair instructions to link tests to potential failures and revealing connections between different symptoms through test outcomes.
18. The system of claim 15 wherein the means for performing reachability analysis determines where failure modes in the fault tree are cross-functionally dependent on more than one symptom.
The system for creating a fault model as described above, where the module that performs reachability analysis determines instances where failure modes in the fault tree are cross-functionally dependent on more than one symptom. This pinpoints scenarios where a single failure can manifest in multiple ways, thus affecting different parts of the system and producing varying symptoms.
19. The system of claim 15 wherein the means for assembling the fault model creates rows of failure modes, creates columns of symptoms, and places a correlation values in intersections of the rows and the columns.
The system for creating a fault model as described above, where the module that assembles the fault model creates rows of failure modes, creates columns of symptoms, and places a correlation value in intersections of the rows and the columns. This table structure visually represents the relationships between failures and symptoms and the strength or probability of those relationships.
20. The system of claim 15 wherein the structured text document is a service procedure document, and the hardware or software system is a vehicle or a vehicle sub-system.
The system for creating a fault model as described above, where the structured text document is a service procedure document and the hardware or software system is a vehicle or vehicle subsystem. This means the system is tailored to automatically build fault models for automotive diagnostics by processing existing repair manuals and diagnostic guides.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
March 10, 2011
September 3, 2013
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.